SPONSORED BY: State of Aadhaar

Page created by Emily Bowman
 
CONTINUE READING
SPONSORED BY: State of Aadhaar
State of Aadhaar Report 2017-18

                                                  1
      MAY 2018                    SPONSORED BY:
SPONSORED BY: State of Aadhaar
SPONSORED BY: State of Aadhaar
STATE OF AADHAAR
REPORT 2017-18
MAY 2018

Ronald Abraham
Elizabeth S. Bennett
Rajesh Bhusal
Shreya Dubey
Qian (Sindy) Li
Akash Pattanayak
Neil Buddy Shah

Sponsored by
SPONSORED BY: State of Aadhaar
SPONSORED BY: State of Aadhaar
About IDinsight

IDinsight's mission is to help policymakers and practitioners use rigorous data and evidence to make more
socially impactful decisions. We carefully tailor a wide range of quantitative tools to help our clients design
better public policies, rigorously test those ideas, and take informed action to improve lives at scale.

IDinsight leads learning partnerships across Asia and Africa, working with governments, multilaterals,
foundations, social businesses, and innovative NGOs. We have worked with more than 70 clients across
17 states in India and more than a dozen countries. We work across a range of sectors, including digital ID,
education, financial access, governance, health, and sanitation.

We have offices in Dakar, Johannesburg, Lusaka, Manila, Nairobi, New Delhi, San Francisco, and Washington,
D.C. To learn more, visit www.IDinsight.org, and follow on Twitter @IDinsight.

About Omidyar Network

Omidyar Network is a philanthropic investment firm. We create opportunity for people to improve their lives
by investing in market-based efforts that catalyse economic and social change. In India, we focus our efforts
on helping the hundreds of millions of lower-income Indians, from the poorest among us to the existing
middle class.

Omidyar Network has committed more than $1 billion to for-profit companies and nonprofit organisations
that foster economic advancement and encourage individual participation across multiple areas, including
Digital Identity, Education, Emerging Tech, Financial Inclusion, Governance & Citizen Engagement, and
Property Rights.

To learn more, visit www.omidyar.com, and follow on Twitter @omidyarnetwork #PositiveReturns.

Suggested citation

Abraham, Ronald, Elizabeth S. Bennett, Rajesh Bhusal, Shreya Dubey, Qian (Sindy) Li, Akash Pattanayak,
and Neil Buddy Shah. State of Aadhaar Report 2017-18. Report. IDinsight, 2018.

Comments or questions

We welcome your feedback on this report. Please write to us with your comments or questions
to StateofAadhaar@IDinsight.org.

Disclaimer

The information contained in this report is prepared by IDinsight and commissioned by Omidyar Network.
It is furnished to the recipient(s) for free distribution and use. The authors have made their best efforts to
ensure the accuracy and completeness of the information in this report but make no representations or
warranties therein and expressly disclaim any liabilities based on such information or on omissions. Each
recipient should therefore conduct her or his own analysis of any information contained in this report.

While Omidyar Network is pleased to sponsor this report, the conclusions, opinions, or points of view
expressed in the report do not necessarily represent the views of Omidyar Network.
SPONSORED BY: State of Aadhaar
SPONSORED BY: State of Aadhaar
PREFACE

In last year’s State of Aadhaar Report, we drafted a comprehensive, empirical, and structured review
of the Aadhaar landscape. This foundational assessment aimed to promote a more fact-based understanding
of the ecosystem at-large. And more critically, we used this analysis to identify areas where more research
was needed to understand whether and how Aadhaar can advance the welfare of India’s residents.

For this year’s report, we aimed to dive deeper into key issues in order to contribute primary research
on critical topics highlighted in last year’s report. After a series of consultations, we decided to conduct
state-representative large-scale surveys to understand how individuals interact with and perceive Aadhaar.
This allowed us to empirically explore important issues that to date were discussed without the support
of rigorous evidence.

The main results of that survey are contained within these pages, supplemented with other independent
research and additional data from the government.

During our initial phase of research design we received valuable inputs from a wide range of thought-leaders
and researchers, including: Alan Gelb, Anurodh Giri, Avani Kapur, Reetika Khera, Vijay Madan, Anit Mukherjee,
Ajay Bhushan Pandey, Vaishnavi Prathap, Malavika Raghavan, Emrys Schoemaker, Ajay Shah, and Janaki
Srinivasan. We are thankful for their time and efforts.

This report would not have been possible without our incredible team at IDinsight. We are grateful to Doug
Johnson for his thoughtful feedback as well as Ruchika Joshi for her contributions in our efforts to engage
in a data-driven discourse. We also appreciate technical input and reviews from Qayam Jetha, Akib Khan,
Jeff McManus, Marc Shotland, and Dan Stein – as well as feedback on our questionnaire from Heather
Lanthorn.

We are thankful to Adi Raval and Divya Sooryakumar for their contributions to our communications strategy.
We also benefited from the support and patience of the entire IDinsight Delhi office as they helped us
brainstorm and work through these important issues.

And we appreciate the efforts of Syed Maqbool, Vinod Sharma, and Lalit Nayak who helped to lead our field
teams in Andhra Pradesh, Rajasthan, and West Bengal, respectively.

Omidyar Network (ON) has continued to be a supportive thought-partner throughout this entire journey.
In particular, we would like to convey our sincere appreciation for the efforts of CV Madhukar, Subhashish
Bhadra, Roopa Kudva, Mike Kubzansky, and ON’s entire Digital ID team.

The report itself has greatly benefited from the expertise of our design firm, Studio Subu, as well as our
editor, Allan R. Gold.

Lastly, for their tireless efforts, we would like to thank our co-authors: Neil Buddy Shah, Qian (Sindy) Li,
Rajesh Bhusal, Shreya Dubey, and Akash Pattanayak.

As always, any shortcomings in the report are our own. We hope this report facilitates meaningful dialogue
on digital identity – both in India and abroad.

Ron and Elizabeth
SPONSORED BY: State of Aadhaar
State of Aadhaar Report 2017-18

8
State of Aadhaar Report 2017-18

            EXECUTIVE SUMMARY
    Aadhaar provides identification to more than 1.2 billion
    Indian residents. Its scale, ability to uniquely identify                                        STATE OF AADHAAR
    individuals, and digital interface make it a compelling                                          SURVEY 2017-18
    identification platform. These same features also raise     RAJASTHAN
    questions about privacy, data security, and exclusion.        965 HH

    The current debate on Aadhaar is binary in nature: either
    Aadhaar should be jettisoned or scaled aggressively.
    However, the facts on the ground—as revealed by our
    three-state survey on Aadhaar, the largest to date—
    support a more nuanced approach. Our report’s key
                                                                                                                WEST BENGAL
    takeaways explain why.                                                                                        840 HH

    IDinsight’s State of Aadhaar initiative aims to catalyse
                                                                                                ANDHRA PRADESH
    data-driven discourse and decision-making in the Aadhaar
                                                                                                   1,142 HH
    ecosystem. This report, the initiative’s flagship output,
    aims to provide a holistic and empirically grounded                                       Note: Numbers indicate randomly
    assessment of the state of Aadhaar.                                                       sampled rural households in each state

          KEY TAKEAWAYS

    Aadhaar’s coverage is widespread,
    but the quality of the data has room for
                                                                    1.5X             more self-reported errors in
                                                                                     Aadhaar compared to voter IDs
                                                                    Demographic error-rate in Aadhaar is 8.8%
    improvement.                                                    vs. 5.7% for voter ID

    A majority of Indians value privacy,
    but also approve of linking Aadhaar
                                                                    87%              rural residents approve
                                                                                     mandatory linking of Aadhaar
                                                                     Approval for mandatory linking of Aadhaar to
    to services.                                                     government services; for private services, corresponding
                                                                     figure is 77%

    Aadhaar's analog version (the letter) is
    much more widely used to open bank
                                                                    17%              bank a/c holders used
                                                                                     Aadhaar e-KYC to open a/c
                                                                     67% used Aadhaar letter; rest used other IDs; data for
    accounts than its digital version (e-KYC).                       most recently opened bank account

    Exclusion from food ration (PDS) due to
    Aadhaar-related factors is significant, but
                                                                     2.2%                of food ration recipients in
                                                                                         Rajasthan excluded monthly
                                                                                         due to Aadhaar-related factors
    lower than non-Aadhaar factors.                                  This extrapolates to ~1.2 million people per month in
                                                                     Rajasthan; non-Aadhaar factors account for 6.5%
                                                                     (or ~ 3.7 million people) of food ration (PDS) recipients
                                                                     excluded per month
Unless specified, all data cited in the executive
summary are from our survey and apply to rural areas
of Andhra Pradesh, Rajasthan, and West Bengal.
AADHAAR PLATFORM
Aadhaar has rapidly become the foundational identity document of Indian residents. It has achieved
near-universality in its coverage and acceptance in India. We review key highlights of what has worked,
what has not, and recommended next steps.

   1.2 BILLION                                                  271 MILLION
   residents currently enrolled on Aadhaar                      unique individuals used their Aadhaar to digitally
   platform with over 90% adult saturation in                   authenticate themselves in Feb 2018 (UIDAI 2018).
   most Indian states (UIDAI 2018)                              Even more use Aadhaar’s analog version (the letter)

1. WHAT HAS WORKED?                                             2. WHAT HAS NOT WORKED?
Most people have an Aadhaar number                              Aadhaar data has more self-reported
and most felt it was easy to enroll                             errors than the voter ID database

We find no evidence of differences in enrolment by gender,
                                                                   ERROR COMPARISION (VOTER ID VS AADHAAR)
caste, religion, or education level.
                                                                                               IN VOTER ID       IN AADHAAR

USER PERCEPTION OF AADHAAR ENROLMENT PROCESS                     ANDHRA PRADESH                2.4%              8.0%

                                                                 RAJASTHAN                     2.6%              4.8%
   AP
                                                                 WEST BENGAL                   10.4%             12.2%
   RJ
                                                                   ERRORS WITHIN AADHAAR
  WB
                                                                                               NAME       ADDRESS        DOB

        0%      20%       40%          60%       80%     100%    ANDHRA PRADESH                3.8%       1.2%           2.6%

             EASY       NEUTRAL              DIFFICULT           RAJASTHAN                     1.5%       0.7%           2.4%

                                                                 WEST BENGAL                   5.6%       1.7%           3.5%
In Andhra Pradesh and West Bengal, a relatively small
number of people were wrongly charged an enrolment fee.         People also encountered more challenges with the process
However, in Rajasthan 24% paid a fee.                           of updating their Aadhaar, compared to the process of
                                                                enrolling, such as paying higher-than-required fees.

3. EMERGING POLICY RECOMMENDATIONS

Facilitate updates to correct Aadhaar data                      Reinforce awareness of fees and
To reduce the number of errors in Aadhaar data, conduct         processes for enrolment and updates
“update campaigns” and set up camps, similar to the             Of those who paid higher-than-required fees, majority
enrolment camps that were well-received and worked              were unaware of the actual cost. Conduct awareness
well in increasing Aadhaar coverage.                            campaign around fees and processes, especially for
                                                                vulnerable groups. Install a strong grievance redress
                                                                mechanism.
AADHAAR AND
    FINANCIAL INCLUSION
The “Jan Dhan” banking scheme, Aadhaar, and mobile (together called “JAM”) are meant to play a pivotal role in the
inclusion of unbanked, and underbanked, into the formal financial sector. There are two main channels through which
Aadhaar can help increase financial inclusion:

     1    Open a bank account with an
          Aadhaar letter or e-KYC                                 2         Increase account usage
                                                                            through microATMs & DBTs

      48 TO 138 MILLION                                            435
      increase in e-KYC verifications from FY                      govt. schemes use DBTs to transfers benefits
      2016-17 to 2017-18 (NPCI 2018)                               to bank accounts (some via Aadhaar),
                                                                   incentivising account usage

1. WHAT HAS WORKED?                                             2. WHAT HAS NOT WORKED?
Aadhaar has become a common                                     Use of microATMs remains low; e-KYC
proof-of-ID for bank account opening                            may not quicken a/c opening

The ubiquity of Aadhaar and its universal acceptance has        Only 17% of account-holders had recently used
meant a high proportion of people use it to open a bank         a microATM.
account. However, 66.9% used their Aadhaar as an analog,        No meaningful difference in time reported to open
paper ID not e-KYC (used by 17.2%).                             an account with e-KYC vs. traditional KYC.

USE OF AADHAAR IN OPENING MOST RECENT
                                                                % OF RESPONDENTS WHO GOT ACCESS TO AN ACCOUNT
BANK ACCOUNT
                                                                WITHIN 1 DAY
                 66.9%                17.2%      15.9%

                                                         100%                                                    39.0%

                                                                                                                37.5%
    USED AADHAAR AS ID

    USED AADHAAR E-KYC                                              DID NOT USE E-KYC

    DID NOT USE AADHAAR                                             USED E-KYC

3. EMERGING POLICY RECOMMENDATIONS

Strengthen business correspondent                               Push toward mobile-based financial
network                                                         services
Too few people currently have access to alternatives to         In the absence of a sustainable BC network, there is
brick-and-mortar banks. Ready access to microATMs could         opportunity for mobile-based financial services to provide
change this; however, the network supporting business           greater access to accounts, and help reduce dormancy.
correspondents (BCs) needs to be strengthened.                  These services are able to use Aadhaar as an identity
                                                                platform.
AADHAAR AND
      FOOD RATION (PDS)
Aadhaar is envisioned as a key reform in India's social protection programmes for the poor, especially the
delivery of food ration subsidies via the Public Distribution System (PDS). Aadhaar's two main uses in PDS
have grown steadily over the last year.

    1    Aadhaar seeding (aims to
         remove duplicate and fake                                        2   Aadhaar-based biometric
                                                                              authentication (aims to prevent
         entries from databases)                                              access to non-genuine persons)

        72% TO 82%                                                            35% TO 57%
        increase in percentage of seeded PDS                                  increase in percentage of PDS shops using
        beneficiary households across all states                              e-PoS devices pan-India (Lok Sabha 2018)
        in India (source in main report)

1. WHAT HAS WORKED?                                                       2. WHAT HAS NOT WORKED?
Households feel secure that no one                                        Exclusion from receiving benefits due to
else can take their ration                                                Aadhaar-related factors significant

Majority in Andhra Pradesh and Rajasthan prefer                           0.8%, 2.2%, and 0.8% of PDS beneficiaries in rural Andhra
Aadhaar-based PDS delivery as they perceive biometric                     Pradesh, Rajasthan, and West Bengal, respectively, are
authentication prevents identity fraud.                                   excluded from their entitlements due to Aadhaar-related
                                                                          factors. This extrapolates to ~2 million individuals monthly.

                                OPINION
                                                                          However, non-Aadhaar reasons, such as ration
                    Worse                 Better                          unavailability, contribute much more to exclusion from PDS.
                                                                                                 9.9%
                            REASONS                                            MONTHLY
                                                                                                        0.6%
                                                                               EXCLUSION
                    No-one can      No-one can                                                          2.2%
                                                                               FROM PDS
                    proxy           take ration                                                         ~1.2M PEOPLE

                    More                   Less                                                         0.7%   6.2%
                    technical         technical                                                                        0.2%
                    issues               issues                                                                               0.8%
                                                                                                                       0.1%   ~0.5M
40%     20%    0%                                  0%   20%   40%   60%                                                       PEOPLE

                            RAJASTHAN
                                                                                                        6.5%
                            ANDHRA PRADESH                                                                             5.2%
                                                                                      0.8%
                                                                               1.1%
                                                                                      ~0.3M
But biometrics is also the reason a quarter of the                                    PEOPLE
beneficiaries consider the system worse, as it reduces                                0.3%

flexibility to send a proxy to collect ration on one’s behalf.                 AP                RJ            WB
                                                                               (AUG-NOV 2017)             (OCT 2017-JAN 2018)
                                                                                            (SEP-DEC 2017)

                                                                                 NON-AADHAAR REASONS
For a more comprehensive treatment of these topics, please
                                                                                 AADHAAR AND NON-AADHAAR REASONS
read the full report. To download the report, the technical
appendix with tables, and learn about our research                               AADHAAR REASONS ONLY
methodology, please visit www.StateofAadhaar.in.                                 UNKNOWN
3. EMERGING POLICY RECOMMENDATIONS                          LEGAL HIGHLIGHTS OF THE
                                                            AADHAAR PLATFORM
                                                            In August 2017, the Supreme Court ruled that
To reduce programme exclusion,
                                                            privacy is a fundamental constitutional right.
re-consider monthly biometric
authentication
                                                            An overwhelming majority of Indians care about
Explore alternative technologies (e.g., offline             their right to privacy; more than 96% of
authentication) or processes (e.g., annual verification)    respondents stated it is important for them to
that provide a similar secure experience, but reduce        know how their Aadhaar information is used by
Aadhaar-related exclusion.                                  the government.

Enforce rules allowing alternate IDs                        AADHAAR AND TELECOM:
or authentication mechanisms                                QUICK HIGHLIGHTS
Mechanisms are in place to ensure benefits are not denied
due to Aadhaar: enforce these strongly. Awareness           77% of respondents approved of the mandatory

of other authentication means (e.g., mobile OTP) is low:    linking of Aadhaar to services, including mobile

popularise them.
                                                            More than 50% of respondents who got a SIM
                                                            card in late 2016 or after, used e-KYC to do so
TABLE OF CONTENTS

Figures                                           1

Introduction                                      2

Aadhaar Platform                                 4
Aadhaar enrolment
Aadhaar usage and awareness
Aadhaar data quality and updating data

Legal                                            13
Right to privacy judgement
Mandatory linking of Aadhaar to services
Data protection law

Financial Inclusion                              17
Opening a bank account
Using a bank account

Social Protection: Food Ration                   22
User experience and exclusion
User perception
Government Savings

Emerging Uses: Spotlight on Telecommunications   29

Looking Ahead                                    30

References                                       32
State of Aadhaar Report 2017-18

FIGURES

Figure 1.1     Significant Aadhaar developments in Financial Year (FY) 2017-18
Figure 1.2     State of Aadhaar Survey 2017-18, sample description

Figure 2.1     Interactions with the Aadhaar system
Figure 2.2     Comparison of SOA survey data and UIDAI data on enrolment
Figure 2.3     Comparison of saturation level in Aadhaar and voter ID for adult population
Figure 2.4     User perception of Aadhaar enrolment process
Figure 2.5     Different uses of Aadhaar by residents
Figure 2.6     Aadhaar authentication and unique residents authenticated over time, Apr 2016 – Feb 2018
Figure 2.7     Proportion of Aadhaar awareness levels among survey respondents
Figure 2.8     Comparison of errors in Aadhaar and voter ID
Figure 2.9     Types of error on Aadhaar letters with a demographic error

Figure 3.1     Importance of knowing how the government uses respondents' data (All three states: Andhra Pradesh,
               Rajasthan, West Bengal)
Figure 3.2     Approval ratings on mandatory linking of Aadhaar to services, by entity type (All three states: Andhra
               Pradesh, Rajasthan, West Bengal)

Figure 4.1     Proportion of respondents who used Aadhaar for bank account openings, since 2014
Figure 4.2     Proportion of respondents who possessed other IDs among those who "Used Aadhaar as ID"
               to open an account, since 2014
Figure 4.3     Monthly number of successful Aadhaar e-KYC verifications as per NPCI, Apr 2016 – Feb 2018
Figure 4.4     Proportion of respondents who got access to account within 1 day among those who used e-KYC
               and those who did not
Figure 4.5     Monthly transactions of AEPS and APBS, Apr 2016 – Feb 2018
Figure 4.6     Proportion of respondents who used their bank accounts by whether they receive DBTs or not

Figure 5.1     Proportion of beneficiaries seeded in India’s major social protection programmes
Figure 5.2     Proportion of fair price shops with e-PoS devices across India, Mar 2017 & Mar 2018
Figure 5.3     Average monthly exclusion from PDS and role of Aadhaar, Aug 2017 – Jan 2018
Figure 5.4     Breakdown of monthly exclusion rate by reasons for exclusion, for Rajasthan’s PDS, Sep 2017 - Dec 2017
Figure 5.5     Number of attempts required for successful Aadhaar-based Biometric Authentication in PDS
Figure 5.6     Opinion on whether use of Aadhaar for PDS delivery is better or worse than previous system
               and associated reasons
Figure 5.7     Government of India's savings estimate from Aadhaar, DBT, & other initiatives

Figure 6.1     Proportion of respondents who used Aadhaar for SIM cards since late 2016
Figure 6.2     Proportion of respondents got SIM activated in 1 day among those who used e-KYC vs. those who
               did not use e-KYC (All three states: Andhra Pradesh, Rajasthan, West Bengal)

                                                                                                                        1
State of Aadhaar Report 2017-18

          INTRODUCTION
    Aadhaar currently provides identification to more than        For the first time in Aadhaar’s history, we now have an
    1.2 billion Indian residents, or more than 90 percent of      opportunity to catalyse such a data-rich discourse. In the
    India’s population. Its scale, ability to uniquely identify   last one year, more robust and independent evidence on
    individuals, and digital interface make it a compelling       how Aadhaar is working has been generated than in any
    identification platform for use by the government, civil      preceding year. The State of Aadhaar Survey 2017-18 (SOA
    society, and private sector. These same features, however,    survey) advances the evidence base with the largest
    raise questions about privacy, data security, and exclusion   survey on Aadhaar to date. The survey provides data
    of vulnerable populations.                                    representative of the rural populations of three states in
                                                                  India: Andhra Pradesh, Rajasthan, and West Bengal. Given
    These concerns came to the fore on multiple occasions         the three states’ varying cultural contexts, governance
    this past year, one of the most eventful since Aadhaar’s      capacity, and take-up of Aadhaar use-cases, they also
    inception in 2009. In Figure 1.1, we provide a timeline       provide an indicative understanding of Aadhaar’s status
    of these significant events.                                  quo nationally. This survey, in conjunction with other
                                                                  published research and official data, form the basis for this
    Throughout this year, however, the discourse was fractious
                                                                  report.
    and polarized. Those challenging Aadhaar contended
    that it is a tool for surveillance and disempowerment         Our findings suggest that important reforms are required
    of the poor and called for jettisoning the project.           for Aadhaar to realise its potential, while simultaneously
    Those in government posited that Aadhaar empowers             limiting the risks it poses. To make genuine progress on
    the poor, saves money for the exchequer, and increases        the role of Aadhaar within India, we invite a vigorous
    state capacity. Their resulting policy prescription is        debate on our findings and their policy implications.
    to aggressively scale the use of Aadhaar.                     We hope this debate advances objective, data-driven
                                                                  discourse and decision-making on Aadhaar, and motivates
    We at the State of Aadhaar initiative believe that instead
                                                                  more independent research to answer questions as they
    of this false dichotomy, India requires a more nuanced
                                                                  arise. This type of empirical approach, we believe, will best
    and empirically grounded discourse and decision-making
                                                                  serve the interests of the country’s more than 1.3 billion
    ecosystem around Aadhaar. We advocate for a careful and
                                                                  residents.
    case-by-case evaluation of Aadhaar’s applications using
    objective data and evidence, in order to inform whether
    a given use should be advanced, adjusted, or abandoned.

2
State of Aadhaar Report 2017-18

    Figure 1.1. Significant Aadhaar developments
    in Financial Year (FY) 2017-18
                                                                                 RESEARCH METHODOLOGY
    June 1: Ministry of Finance order mandates linking                           The State of Aadhaar Report 2017-18 combines primary
    of Aadhaar with bank accounts
                                                                                 data collected from the SOA survey with secondary
    June 9: Supreme Court allows Aadhaar-PAN linkage                             research that draws on administrative data and published
    August 24: Supreme Court upholds the right to privacy                        research on Aadhaar.
    as a fundamental right

    October 16: Newspaper reports claim a 11-year old dies
    of starvation as her family was denied food subsidies                        State of Aadhaar Survey 2017-18
    due to their ration card not being linked to Aadhaar
                                                                                 The survey is designed to provide representative state-
    November 19: Unique Identification Authority of India
    (UIDAI), in a Right to Information (RTI) reply, stated                       level estimates of the rural populations of Andhra Pradesh,
    that 210 government websites disclosed Aadhaar data                          Rajasthan, and West Bengal. The states were chosen to be
    November 27: Government of India releases a white                            geographically distributed across India, representing
    paper on “Data Protection Framework for India”
                                                                                 diverse contexts of culture and state capacity. In addition,
    January 4: Journalist for The Tribune reported that she                      these states have varying penetration of Aadhaar in terms
    was able to buy unauthorised access to a UIDAI portal
    enabling her to get demographic details of any individual                    of enrolment and linkages to services.
    upon entering a valid Aadhaar number
                                                                                 A total of 2,947 households were surveyed in person (see
    January 10: UIDAI announced virtual ID, limited e-KYC,
    and UID token                                                                Figure 1.2). The sample size was selected to maximise the
                                                                                 precision of the estimates while managing operational
    January 17: Five-judge constitution bench begins
    hearing 30 petitions challenging Aadhaar                                     constraints. We used the Election Commission's public,
                                                                                 online voter ID database to construct the sampling frame.
    Sources: Supreme Court orders, MeitY notification,
                                                                                 Sample selection used the probability proportional to size
    newspaper reports
                                                                                 (PPS) method to sample districts (stratified by region)
                                                                                 and sub-district units. Simple random sampling was used
                                                                                 to select households and again to select individual
                                                                                 respondents within households.1
     New to Aadhaar?
     This report is written assuming a basic                                     Secondary Research
     understanding of Aadhaar and its uses.
                                                                                 Other sources for this report include data from various
     If you require a primer, however, we
                                                                                 public portals maintained by the central and state
     suggest reading our first report:                                           governments, government reports and documents, and
     State of Aadhaar Report 2016-17.                                            published independent research. Please see our full list
                                                                                 of references at the end of the report.

      Figure 1.2: State of Aadhaar Survey 2017-18, sample description

                                                                                AP             RJ              WB             ALL 3 STATES

        Number of sampled households (and respondents)                          1,142          965             840            2,947

        Number of sampled household members                                     4,454          5,430           3,785          13,669

        Number of sampled districts                                             6 (of 13)      8 (of 33)       7 (of 23)      21 (of 69)

        Dates of survey                                                         Nov-Dec        Dec 2017 -      Jan-Feb        Nov 2017 -
                                                                                2017           Jan 2018        2018           Feb 2018

        2017 rural population                                                   34.8m          57.7m           65.9m          158.4m

      Note: 2017 rural population estimates calculated using data from UIDAI and Census.
      A detailed technical note on the survey methodology is available on www.StateofAadhaar.in.

1. One adult per household was selected to be the respondent. Some survey questions only concern the respondent themselves, while others concern all members
   of the household, including children.                                                                                                                       3
State of Aadhaar Report 2017-18

          AADHAAR
          PLATFORM
    Aadhaar is the world’s largest publicly run digital identity
    database and it continues to grow. The database contains       Figure 2.1: Interactions with the Aadhaar System
    1.2 billion enrollees and has facilitated nearly 24 billion
    digital authentications (Unique Identification Authority of
    India 2018). This demonstrates phenomenal growth since
    the first Aadhaar was issued in 2010, yet key questions             ENROL                     USE
    related to the platform – enrolment and updating
    processes, usage, and data quality – remain unanswered.
    As we show in Figure 2.1, these areas comprise the major
                                                                                      UPDATE PROCESS
    touchpoints with the Aadhaar system.

    With the State of Aadhaar Survey 2017-18 (SOA survey),                                               UPDATE
                                                                            FIX ERRORS
    along with government administrative data, we begin                        IF ANY
                                                                                                         INFO AS
                                                                                                        REQUIRED
    providing initial answers in this section. Specifically,
    we seek to answer the following critical questions:

4
State of Aadhaar Report 2017-18

1.   What are people’s experience with the Aadhaar                                       An important motivation for Aadhaar at the time of its
     enrolment process? To what extent are people                                        inception was to provide formal identification to those
     excluded from the Aadhaar platform itself?                                          who did not have one. However, in our survey, more
                                                                                         than 99 percent of respondents in all three states had
2. How do people use their Aadhaar? What are the most                                    at least one other form of ID at the time of enrolment.5
     common usages?                                                                      This finding is also in line with an RTI response that
                                                                                         demonstrated the number of enrolments that used
3. What is the quality of the demographic data captured
                                                                                         the introducer system accounted for only 0.02 percent
     by Aadhaar? What are individuals’ experiences
                                                                                         of total enrolments (Sharma (RTI) 2015).6
     regarding fixing errors or updating their Aadhaar?

The next subsections discuss our findings related to
                                                                                              Figure 2.2: Comparison of SOA data and UIDAI
these questions. This is followed by a discussion on our
                                                                                              data on enrolment
recommended next steps informed by the findings.
                                                                                                         % OF AADHAAR ENROLMENT

                                                                                                                                                           97.6%
                                                                                                 AP                                                       92.6%

AADHAAR ENROLMENT                                                                                 RJ
                                                                                                                                                    87.5%
                                                                                                                                                    86.7%

Overall Enrolment Trends in the                                                                 WB
                                                                                                                                                          93.2%
                                                                                                                                                          94.0%
SOA survey
                                                                                                              SOA DATA                    UIDAI DATA
The enrolment saturation rates from the SOA survey
                                                                                                Note: Error bars denote 95% confidence intervals
support the UIDAI’s saturation data (see Figure 2.2).
                                                                                                Number of observations: AP (4,448); RJ (5,396); WB (3,778)
Beyond the high saturation rate of more than 90 percent,                                        Data sources: State of Aadhaar Survey 2017-18;
                                                                                                UIDAI State-wise Saturation Report, March 31, 2018
we looked into enrolment by different demographics.
We do not find that members of vulnerable communities
are less likely to be enrolled.2 This is indicative that
exclusion from Aadhaar is not only low, but also                                         The most common forms of identification were ration
not systematically biased against the poor or vulnerable.                                cards and voter IDs. While both of these IDs are widely
                                                                                         accepted, it should be noted that they also have
For those who tried to enrol in the Aadhaar system                                       limitations. Ration cards, in most states, are provided
but were turned away (1 percent of all household members                                 at the household level, not the individual level. Voter IDs
across the three states), we sought to understand why.                                   are restricted to those eligible to vote—meaning eighteen
The most common reason for rejection was "ineligibility                                  years old and above. Thus, neither form of ID is universal
due to biometrics" which was 30.9 percent across all three                               in quite the same way as Aadhaar. A comparison of
states.3,4 Given the centrality of biometric capture to the                              saturation rates for Aadhaar and voter ID in the three
Aadhaar platform, the UIDAI should look into this issue                                  states we surveyed can be seen in Figure 2.3.
further to determine root causes and to develop
technological or manual solutions.

2. We tested for differences in the following populations: gender, religion, age, and caste. We found that respondents above the age of 60 are more likely to be
   enrolled. When conducting hypothesis tests for difference in outcomes between various groups we correct for multiple hypothesis testing using the Bonferroni
   correction. This applies to all hypothesis tests contained in this report.

3. “Ineligibility due to biometrics” refers to cases where residents were told that their biometrics are not of sufficient quality to enrol in Aadhaar.

4. There was variation across states; however, the number of state-level observations was too small to be able to report on any differences across states.

5. In the survey, we used voter rolls to construct the sampling framework. We are fairly confident that our sample captures households where some members of
   the household may have been left off the voter roll (as evidenced by the fact that more than 10 percent of adults within the sample did not have a voter ID).
   However, one limitation of the survey is that we would have been unable to capture households where no single member is on the voter list. For more details
   about the survey methodology, including limitations, please see www.StateofAadhaar.in.

6. The introducer system is intended to be an avenue of enrolment for individuals unable to provide proofs-of-identity.                                            5
State of Aadhaar Report 2017-18

                                                                                          highest incidence of deviation was found in Rajasthan,
        Figure 2.3: Comparison of saturation level in                                     with 23.7 percent of people stating that they paid to enrol
        Aadhaar and voter ID for adult population                                         in Aadhaar. Those in Andhra Pradesh and West Bengal
                                                                                          experienced this issue as well with 9.0 percent and 5.7
                   PROPORTION OF RESPONDENTS
                                                                                          percent of people, respectively, paying to enrol. Those who
                                                                   99.7%                  enrolled in an Aadhaar camp were less likely to have paid
           AP                                                   89.5%
                                                                                          for enrolment in all three states.
                                                                  98.0%
            RJ                                                 82.7%                      While many enrollees reported facing inconveniences such
                                                                                          as failed enrolment or the charging of unnecessary fees,
                                                                  97.8%
          WB                                                     93.7%                    an overwhelming majority of the respondents in all three
                                                                                          states indicated they found the enrolment process to be
                          AADHAAR                    VOTER ID
                                                                                          “easy” (see Figure 2.4).
          Note: Error bars denote 95% confidence intervals
          Number of observations (Aadhaar): AP (3,415); RJ (3,553);
          WB (2,810) Number of observations (voter ID): AP (3,399);
          RJ (3,553); WB (2,807)

                                                                                          AADHAAR USAGE AND
          Data source: State of Aadhaar Survey 2017-18

    Analysing the enrolment process
                                                                                          AWARENESS
    The UIDAI has built a sizeable infrastructure for the                                 A critical but largely unanswered question is how do
    purposes of enrolment. We describe this infrastructure in                             people tend to use their Aadhaar. Below, we explore the
    the Aadhaar Architecture chapter of the State of Aadhaar                              most common types of uses and look into how many
    Report 2016-17. In this year’s report we focus on the                                 people use each type.
    efficacy of these systems, the resulting user experience,
    and user perceptions of the enrolment process.
                                                                                          Usage to establish identity
    UIDAI regulations state that enrolment in Aadhaar is free
                                                                                          The most common reported use of Aadhaar for
    and no entity can charge a fee for this service (UIDAI
                                                                                          establishing identity is providing a photocopy of one’s
    2016). However, findings from our survey show that this
                                                                                          Aadhaar letter.7 More than 93 percent of respondents in all
    regulation has not always borne out in practice. The
                                                                                          three states claimed they have used Aadhaar in this way to
                                                                                          establish their identity (see Figure 2.5). The next most
                                                                                          common usage is fingerprint authentication. While this use
        Figure 2.4: User perception of Aadhaar
                                                                                          is widespread in Andhra Pradesh (95.4 percent) and
        enrolment process
                                                                                          Rajasthan (76.7 percent), it is less common in West Bengal
                   PROPORTION OF RESPONDENTS                                              (48.8 percent). One probable explanation is the fact that
                                                                     84.7%                West Bengal’s ration system does not currently use
           AP            6.3%                                                             Aadhaar-based biometric authentication at a large scale.
                            9.0%
                                                                                          Iris authentication usage in Andhra Pradesh is much higher
                                                                      84.4%               than in Rajasthan and West Bengal. This is likely due to the
            RJ             5.7%
                                                                                          use of iris authentication in the Public Distribution System
                           9.8%
                                                                                          (PDS) and cash withdrawals from iris-enabled microATMs.
                                                                 77.3%
          WB                  11.1%                                                       Data from the UIDAI shows that the number of overall
                             11.6%
                                                                                          Aadhaar authentications is growing steadily. Similarly, the
                   EASY               NEUTRAL             DIFFICULT                       number of unique IDs (UIDs) authenticated continues to
                                                                                          trend upward (UIDAI 2018). In February 2018, 271 million
          Note: Error bars denote 95% confidence intervals
          Number of observations: AP (1,140); RJ (950); WB (826)                          individuals used their Aadhaar to authenticate themselves,
          Data sources: State of Aadhaar Survey 2017-18                                   representing nearly one in four people in possession of an
                                                                                          Aadhaar (see Figure 2.6).

    7. The Aadhaar letter – sometimes referred to as an “Aadhaar card” – is issued by the UIDAI. The letter contains one’s Aadhaar number, demographic details, and
       photograph. Throughout this report we refer to the letter as an “analog” or “non-digital” ID to differentiate between the biometrically-enabled aspects of the ID.

6
State of Aadhaar Report 2017-18

     Figure 2.5: Different uses of Aadhaar by residents

                   PROVIDED A PHOTOCOPY OF MY AADHAAR CARD

                                                                                                                            93.4%
                                                                                                                              96.8%
                                                                                                                            93.9%

                   SHOWED MY AADHAAR CARD AS A FORM OF ID
                                                                                                                     84.7%
                                                                            39.9%
                                                          24.5 %

                   USED FINGERPRINT AUTHENTICATION
                                                                                                                             95.4%
                                                                                                             76.7%
                                                                                          48.8%

                   USED IRIS AUTHENTICATION
                                                                   26.4%
                       1%
                         2.2%
                                                                                                       ANDHRA PRADESH

                   USED ONE-TIME-PASSWORD AUTHENTICATION                                               RAJASTHAN

                            2.0%                                                                       WEST BENGAL
                          2.0%
                                    9.3%                                                           PROPORTION OF RESPONDENTS

        Note: Error bars denote 95% confidence intervals
        Number of observations: AP (1,142); RJ (952); WB (825)
        Data source: State of Aadhaar Survey 2017-18

     Figure 2.6: Aadhaar authentication and unique residents authenticated over time, Apr 2016 – Feb 2018

              NUMBER OF AUTHENTICATIONS
                                                                                                     1,463
              NUMBER OF UIDS AUTHENTICATED

                                                                                                                     1,094
                                                                                         966
            NUMBERS
            IN MILLIONS

                                                          565          464
                                           391
                           217
                                                                                                      374
                                                                                                                      271
                                                                                         218
                                           143             158             125
                            83

         APR ‘16         JUN ‘16        OCT ‘16         JAN ‘17       APR ‘17          JUN ‘17      OCT ‘17      JAN ‘18         APR ‘18

     Data source: UIDAI Dashboard

Awareness                                                                        See Figure 2.7 for an assessment of awareness levels
                                                                                 for the three states in our survey. The UIDAI has set up
In our survey we asked a series of questions about                               regulations and systems to improve access and also
awareness to understand which features and                                       safeguard its end users. However, lack of user awareness
functionalities people understood regarding Aadhaar.                             has limited the impact of these steps.

                                                                                                                                            7
State of Aadhaar Report 2017-18

8
State of Aadhaar Report 2017-18

EVOLVING FEATURES OF AADHAAR
To try and address growing concerns about privacy and data security, the UIDAI
recently launched three features: Virtual ID, Limited KYC, and UID token.

Virtual ID (VID) is a temporary 16-digit random number that can be generated
by residents against their Aadhaar numbers. VID, along with one’s biometrics,
can be used to authenticate oneself. The introduction of VID tackles privacy
concerns in two ways. One, it prevents agencies from linking databases using
Aadhaar’s unique identifier since each agency will only have access to the
temporary VID. Two, it allows an individual to have a choice about when (and
with whom) they share their Aadhaar number.

Limited Know-Your-Customer (KYC) is being introduced to regulate the storing
of Aadhaar numbers in different databases. As per the UIDAI’s notification, the
authority plans to allow only authentication agencies that are required by law to
receive full demographic information along with the Aadhaar number of an
individual. The remaining agencies will be subject to Limited KYC and will not
be allowed to store Aadhaar numbers upon authentication.

To facilitate uniqueness and security of beneficiaries in the authentication
agencies’ databases, the UIDAI aims to introduce a system called UID Token.
With this system, each individual Aadhaar number is given a unique token ID for
an authentication agency. That token is used for each transaction with that
agency, but not used anywhere beyond that agency (UIDAI 2018). This too
prevents different agencies from linking databases as each agency has a unique
token for an individual that cannot be mapped to the token of that individual
from another database.

Efforts to advance and evolve security features to better protect residents
should be encouraged. However, given low levels of awareness of other security
features such as biometric locking, simply offering these features is unlikely to
result in sufficient uptake (see Figure 2.7). The government should engage in
thoughtful campaigns, targeted toward individual users and institutions
requiring Aadhaar for identification to ensure the objectives of these initiatives
are met. Finally, these measures, and the processes accompanying them, need
careful and rapid testing to ensure they are optimised to provide a seamless
and effective layer of protection.

                                                                                     9
State of Aadhaar Report 2017-18

               Figure 2.7: Proportion of Aadhaar awareness levels among survey respondents

                   CATEGORY                      ISSUE                                                        AP           RJ          WB           ALL 3 STATES

                   Authentication                Awareness of fingerprint authentication                      97.8         87.6        71.5         85.1
                   Awareness
                                                 Awareness of iris authentication                             45.7         15.4        30.8         31.0

                                                 Awareness of OTP authentication                              6.1          10.0        13.9         10.1

                                                 Awareness of all three                                       5.8          7.2         8.7          7.2

t
                   UIDAI                         Enrolment is free of cost                                    40.2         41.0        43.3         41.6
                   Regulations
                                                 Updating should cost only ₹15                                4.8          11.8        8.0          7.3

                   Security Features             Aware of biometric locking & unlocking                       2.0          3.4         3.9          3.1

               Note: The fee charged for updates was ₹15 but has recently been increased to ₹25
               Number of observations: The number of observations reported above varies by question and by state. Most questions were asked to all respondents
               who have an Aadhaar, with variation arising from the removal of “don’t know” responses. However, the question on cost of updating was only asked
               to a relevant subset (275). See the technical appendix on www.StateofAadhaar.in for more details.
               Data source: State of Aadhaar Survey 2017-18

                                                                                               To interpret the errors compared with the most widely
         AADHAAR DATA QUALITY                                                                  used alternative, we also asked if respondents had errors
                                                                                               in their voter ID cards. We see that in Andhra Pradesh and
         AND UPDATING DATA                                                                     Rajasthan the proportion of respondents with errors in
                                                                                               Aadhaar are more than three times and nearly two times
         While saturation levels are important, it is equally
                                                                                               higher than for voter IDs, respectively, while in West
         imperative to ensure that the information collected is
                                                                                               Bengal the Aadhaar error rate is higher, though the
         accurate. Low data quality would lead to errors in seeding
                                                                                               difference is not as extreme (see Figure 2.8).
         and a less reliable e-KYC process. Through our survey, we
         wanted to gain a better understanding of the quality of                               We also wanted to identify which types of information
         the demographic data captured through Aadhaar. To                                     are most error-prone. The most common source of error
         understand this vital component of Aadhaar, we examined                               in Andhra Pradesh and West Bengal was the name of the
         several related indicators, such as the proportion of                                 respondent, while in Rajasthan date of birth contained the
         respondents that self-reported an error in their Aadhaar                              highest number of self-reported errors (see Figure 2.9).
         data, the frequency of updating (due to errors or a need to                           The magnitude of error(s) also matters in terms of how
         update information), and the proportion of duplicates in                              the error affects people. Of those who reported an error
         the system. We also wanted to understand people’s                                     in their name, about one-third claimed their full name
         experience with the updating process.                                                 was wrong while two-thirds noted an incorrect spelling.
                                                                                               We also asked respondents about the perceived source

         Data quality                                                                          of the error. Eighty-nine percent reported their error(s)
                                                                                               was due to data entry mistakes while 10.3 percent stated
         At the peak of the enrolment process, the UIDAI was                                   a pre-existing error in a source document. The remaining
         facilitating enrolment of nearly one million people a day                             errors were attributed to mistakes made by the
         (State of Aadhaar Report 2016-17). This raises the question                           respondents themselves.
         about whether the scale of the enrolment made the
         exercise vulnerable to data entry errors. When asked,                                 The UIDAI has established processes for the correction

         8.8 percent of people stated that their Aadhaar letter                                of errors in Aadhaar. Most who tried to get their error

         contained an error. The errors are self-reported and there                            rectified claimed to be successful. However, only 53.0

         is reason to believe these are a lower-bound estimate for                             percent of those with errors actually reported trying to get

         errors.   8                                                                           the error corrected. An important concern that emerges
                                                                                               from the data on the enrolment process is that of the

         8. We believe self-reported errors will be lower than externally verified errors. This assumes that respondents are more likely to report an error that has affected
            their ability to use their Aadhaar or access services, and less likely to report an error if it does not impede any uses. For example, a name may be misspelled or
            wrongly transliterated from another language into English, but a respondent would likely only report the error if it has led to service denials or other issues.
    10      Additionally, respondents who are not literate might not be able to tell if some information is incorrect.
State of Aadhaar Report 2017-18

respondents who paid to get the error fixed (82.4                                 Updating data
percent), 96.0 percent reported paying more than
the designated fee of ₹25.9                                                       Updating Aadhaar data will become essential to ensuring
                                                                                  accuracy of the information. This primarily involves
In contrast to the high number of respondents reporting                           updating contact details, photographs, and biometric
ease in the enrolment process, the proportion is lower for                        information of children once they turn five and again at
the correction of errors. In Andhra Pradesh and Rajasthan                         age fifteen, per the UIDAI requirements. Some updates are
69.1 percent and 67.0 percent, respectively, stated that                          required on an ad hoc basis (updating one’s address after
they found the process “easy.”10 In West Bengal, 47.8                             a move) while some updates are cyclical (updating one’s
percent of those who attempted to fix errors felt the                             photo every ten years). The proportion of people that
process was “easy.” This indicates that there are challenges                      updated their data in cases where their address had
or barriers in this process that should be addressed to                           changed was low (13.0 percent) amongst the three states.
improve the user experience and to ensure the accuracy                            The proportion was higher (25.5 percent) in the case of
of the Aadhaar data.                                                              updating mobile phones numbers.

    Figure 2.8: Comparison of errors in Aadhaar
    and voter ID

              PROPORTION OF RESPONDENTS
                                                                                  DISCUSSION
                                                     8.0%
       AP                     2.4%                                                As noted above, people were generally satisfied with the
                                                                                  Aadhaar enrolment process. However, they faced greater
                                     4.8%
       RJ                    2.6%
                                                                                  challenges when fixing mistakes or updating information.
                                                                 12.2%            Updates will be required as long as the system is in place.
     WB                                                      10.4%
                                                                                  Consequently, the UIDAI should examine the successes of
                                                                                  the enrolment system and think critically about how to
                   ERROR RATE IN AADHAAR                                          improve the updating process.
                   ERROR RATE IN VOTER ID
                                                                                  As the administrator of the world’s largest biometric
   Note: Error bars denote 95% confidence intervals
                                                                                  database, the UIDAI will face significant challenges
   Number of observations (Aadhaar): AP (4,275); RJ (4,669); WB (3,435)
   Number of observations (voter ID): AP (2,999); RJ (2,934); WB (2,611)          ensuring Aadhaar data remains current and updated.
   Data source: State of Aadhaar Survey 2017-18
                                                                                  A system containing errors is likely to cause problems,
                                                                                  ranging from minor inconveniences to serious threats of
                                                                                  exclusion.
    Figure 2.9: Types of error on Aadhaar letters
    with a demographic error                                                      Similarly, some uses of Aadhaar authentication remain low,
                                                                                  particularly OTP and iris authentication. Given fairly high
              PROPORTION OF RESPONDENTS                                           rates of mobile penetration, it may be worthwhile for the
                                                              48.3%               UIDAI to invest in awareness campaigns aimed at the ease
       AP                                   15.3%                                 of OTP authentication. This could also provide relief for
                                              33.8%
                                                                                  individuals who struggle to authenticate using their
                                              32.2%                               biometrics. Additionally, the overwhelmingly high use of
       RJ                                15.6%                                    Aadhaar in its analog form necessitates a greater focus on
                                                              50.4%
                                                                                  adding security features to the physical copy of Aadhaar
                                                              47.7%               as well. As Aadhaar’s use as an identity platform continues
     WB                                14.5%                                      to increase, it is vital for the underlying quality of its
                                                  30.2%
                                                                                  database to improve, along with well-established
            NAME             ADDRESS              DATE OF BIRTH                   processes to continuously stay up to date. Having achieved
                                                                                  exemplary scale, a renewed focus toward quality and
      Note: Error bars denote 95% confidence intervals
                                                                                  updating processes may be important next steps for the
      Number of observations: AP (333); RJ (223); WB (410)
      Data source: State of Aadhaar Survey 2017-18                                UIDAI.

9. The fee charged for updates was ₹15 but has recently been increased to ₹25.
                                                                                                                                                11
10. The number of respondents who answered this question was less than fifty in Andhra Pradesh and Rajasthan.
State of Aadhaar Report 2017-18

            DUPLICATE AADHAAR "CARDS"
            One of the strongest arguments in favour of Aadhaar is the ability to create a
            database free of duplicate entries. The unique 12-digit number combined with
            individuals’ unique biometric information is meant to create a system in which
            each individual is only entered once.

            Our survey data suggests this is mostly true; however, we found three cases in
            which individuals possessed two Aadhaar letters—containing matching name
            and demographics but with a distinct unique ID number. At 0.1 percent of our
            sample, this represents a small but meaningful number.11

            Based on UIDAI’s performance data, the false acceptance rate was expected to
            be 0.035 percent (State of Aadhaar Report 2016-17). However, given different
            estimation methods and the low incidence in our sample, we cannot strictly
            compare these two statistics.

     11.
     11. A direct extrapolation of this number would be equivalent to 71,300 duplicates across the rural population of all three states. The 95 percent confidence interval
        around this estimate ranges from 3 to 213,900, meaning there is a wide variance on our estimate for how many duplicates may exist. However, we may want to
        think of this number as a lower-bound considering that anyone who intentionally obtained a second Aadhaar (for nefarious purposes) would likely not have
        shared this information with our survey team. Note that we confirmed that these were indeed duplicate Aadhaars and not the respondent mistaking their
        temporary enrollment ID to an Aadhaar number.
12
State of Aadhaar Report 2017-18

       LEGAL
As we write this report, a five-judge constitution bench of   India has also set in motion a consultative process
the Supreme Court of India is hearing 30 petitions            regarding the framing of a much-needed data protection
challenging the legality of Aadhaar. The Court is expected    law (Ministry of Electronics and Information Technology
to answer foundational questions about the digital identify   [MeitY] 2017).
platform. Is the 2016 Aadhaar Act, passed as a Money Bill,
constitutional? Is Aadhaar’s use of biometrics in             In this chapter, we provide a brief overview of these legal
conformity with our newly codified right to privacy? Are      developments in the Aadhaar ecosystem over the last year.
the burdens borne by Indian residents due to the use of       We do not discuss the current set of petitions being heard
Aadhaar justified by the potential gains?                     in the Supreme Court as the matter has not concluded at
                                                              the time of sending this report to print.
In answering these questions, the Supreme Court can play
a crucial role in shaping our institutions and ability to
manage risks in a fast-paced digital world. In August 2017,
its landmark judgement on the right to privacy served as
confirmation of this role. In addition, the Government of

                                                                                                                            13
State of Aadhaar Report 2017-18

     RIGHT TO PRIVACY                                                                        Figure 3.1: Importance of knowing how
                                                                                             the government uses respondents' data
     JUDGEMENT                                                                               (All three states: AP, RJ, WB)

                                                                                                  PROPORTION OF RESPONDENTS
     A nine-judge bench of the Supreme Court of India passed
     a landmark judgement recognizing our fundamental right                                                                                           96.5%

     to privacy as emanating from Article 21 (right to life and                                                                                       96.5%
     personal liberty) and other rights guaranteed under Part III
                                                                                                                                                      97.0%
     of India’s Constitution. The judgement came in light of a
     series of individual petitions linked to Aadhaar combined
                                                                                                       BASIC PERSONAL INFO (NAME, AGE, ETC.)
     with the writ petition filed by retired High Court Justice
                                                                                                       BIOMETRIC INFO (FINGERPRINTS, ETC.)
     K.S. Puttaswamy challenging various aspects of the
                                                                                                       AADHAAR NUMBER
     Aadhaar project, especially its potential violations of
     individual privacy.                                                                      Note: Error bars denote 95% confidence intervals
                                                                                              Number of observations (for basic personal information):
                                                                                              AP (1,135); RJ (939); WB (836)
     Whether India’s constitution allows a right to privacy
                                                                                              Number of observations (for biometric information): AP (1,137);
     remained unsettled before this judgement. While there                                    RJ (932); WB (829)
                                                                                              Number of observations (for Aadhaar number): AP (1,129); RJ
     were two- or three-judge benches that ruled that it was a
                                                                                              (924); WB (820)
     right (Gobind vs State of Madhya Pradesh 1975; Malak                                     Data source: State of Aadhaar Survey 2017-18
     Singh 1980), the Union of India argued against a blanket
     right to privacy, citing larger Supreme Court benches
     (M P Sharma 1954; Kharak Singh 1962).

     Privacy was deemed an undeveloped and elitist concept in                            MANDATORY LINKING OF
     the proceedings (Justice K S Puttaswamy (Retd.) and Anr
     versus Union of India and Ors 2017). Findings from the
                                                                                         AADHAAR TO SERVICES
     SOA survey do not support this view. To a question asking
                                                                                         Using Section 7 of the Aadhaar Act 2016, central and state
     SOA survey respondents whether they consider it
                                                                                         governments have required beneficiaries of 252 schemes
     important (or not) to know how the government uses their
                                                                                         (as of December 2017) to authenticate themselves using
     information, more than 96 percent of respondents in the
                                                                                         Aadhaar (Rajya Sabha Question 2018).
     three surveyed states contended that it was indeed
     important, for all three types of information shown in                              One of these uses, put forward by the Department of
     Figure 3.1. These findings are in line with qualitative                             Telecommunications in March 2016, asks for Aadhaar-
     interviews conducted with low-income individuals in a                               based verification of all existing mobile phone
     collaborative effort between Dalberg, CGAP, and Dvara                               subscribers.12 Similarly, an order in June 2017 by the
     Research. One of their key insights was that respondents                            Supreme Court upheld a move allowing the linking of
     were “highly protective” of their personal information,                             Aadhaar to each individual’s PAN (Binoy Viswam versus
     including their Aadhaar numbers (Privacy on the Line                                Union of India & Ors 2017).
     2017).
                                                                                         We asked respondents in the SOA survey whether they
     The dimensions of privacy discussed in the judgement                                approved or disapproved of the decision to mandatorily
     include an individual's right to seclusion, autonomy over                           link Aadhaar to services offered by the government and
     fundamental personal choices, and the right to control                              the private sector. Approximately 87 percent of people
     dissemination of personal information. Similar to other                             approved of the government’s mandatory use of Aadhaar,
     fundamental rights, the judgement maintained that the                               while 76.9 percent approved Aadhaar’s mandatory use by
     right to privacy can be subject to reasonable restrictions                          the private sector. See Figure 3.2.
     where such an invasion must be justified by law and a
     legitimate state aim, including preserving scarce resources.

     12. The Supreme Court order that led to this notification, however, was more tentative in language on what types of mobile subscribers should be covered and
         whether Aadhaar should be used exclusively in the verification process (State of Aadhaar Blog 2017).

14
State of Aadhaar Report 2017-18

DATA PROTECTION LAW                                                Figure 3.2: Approval ratings on mandatory linking
                                                                   of Aadhaar to services, by entity type (All three
                                                                   states: Andhra Pradesh, Rajasthan, West Bengal)
With an ever-growing digital economy and the right to
privacy judgement, discussion about the need for a data                PROPORTION OF RESPONDENTS

protection law intensified this past year. The MeitY                   APPROVE

constituted an expert committee led by Justice B.N.                                                                       86.5%
                                                                                                                      76.9%
Srikrishna to identify data protection issues in India and
                                                                       NEUTRAL
draft a bill addressing these concerns. The committee
                                                                                6.5%
released a white paper in November 2017 to seek public                            10.1%
views to define the contours of a data protection law in
                                                                       DISAPPROVE
India. The paper outlines seven principles for data
                                                                              7.0%
protection in India—technology agnosticism, holistic                              13.0%
application, informed consent, data minimisation,
controller accountability, structured enforcement, and                      GOVERNMENT
deterrent penalties (MeitY 2017). It draws on existing                      PRIVATE COMPANIES
international laws—such as the European Union’s General
Data Protection Regulation (2016) as well as South Africa’s          Note: Error bars denote 95% confidence intervals
                                                                     Number of observations (for government services): AP (1,136);
Protection of Personal Information Act (2013)—and raises
                                                                     RJ (945); WB (828)
questions on subject matters ranging from potential                  Number of observations (for private services): AP (1,103);
                                                                     RJ (935); WB (823)
definition of different types of data, the territorial scope,
                                                                     Data source: State of Aadhaar Survey 2017-18
extent of exemptions, existence of a data protection
authority, and the model of regulatory approach that must
be adopted.

                                                                organisation, raises potential conflicts, and is not ideal
                                                                given the complexity of regulatory requirements.

DISCUSSION                                                      With the past and forthcoming Supreme Court judgements
                                                                and the establishment of the Srikrishna Committee to
                                                                actively look into data protection, the government has the
The Supreme Court judgement on privacy, the mandatory
                                                                opportunity to institute a strong data protection law and
linking of services, and the forthcoming data protection
                                                                regulator to manage risks posed by Aadhaar and other
law all pivot around the central question of whether and
                                                                digital tools, an opportunity we hope it seizes.
how Aadhaar can be a tool for infringement of privacy.

We believe that Aadhaar, if unregulated, can be a tool that
erodes our fundamental right to privacy. However, it is only
a tool and other public and private digital platforms are
also capable of invading our privacy. Therefore, as we do
with any tool in the public domain, we need to take
advantage of its benefits while managing its risks.

More fundamentally, the threat of privacy infringement
highlights the need for a strong data protection law
backed by an independent, competent, and fully
authorised data protection regulator to enforce it.
Currently, UIDAI acts as both a data controller of Aadhaar
demographic and biometric information and regulator to
ensure the appropriate use of that information. This can be
ineffective given the dual purpose for a single

                                                                                                                                     15
State of Aadhaar Report 2017-18

     AADHAAR DATA DISCLOSURE
     While there has been no unauthorised data disclosure of enrolment data within
     the UIDAI’s Central Identities Data Repository, Aadhaar numbers and
     demographic details have been disclosed on other public portals.

     A report released by the Centre for Internet and Society reported that Aadhaar
     numbers and demographic details of 135 million residents were disclosed on
     four government portals due to lack of stringent IT measures (Sinha and Kodali
     2017). The MeitY further reported that, as per the UIDAI, 210 government
     websites had revealed details of Aadhaar and other personal data (Lok Sabha
     Question 2018).

     Name, address, bank account details, and Aadhaar numbers of more than a
     million beneficiaries of an old age pension scheme in Jharkhand were published
     on a website maintained by the Jharkhand Directorate of Social Security in July
     2017 (Sethi, Bansal, and Saurav 2017).

     A journalist for the Tribune reported a data breach in which she was able to buy
     access to a UIDAI portal. With this access, one could enter a person’s Aadhaar
     number and obtain personal demographic data (Khaira 2018).

16
You can also read